Date: December 17, 2013
Time: 5:00 pm
Location: Meeting room D4.3.106 (Campus map) WU, Welthandelsplatz 1, Building D4, 3^rd Floor

Dear fellow researchers, students and visitors!

You are cordially invited to a research talk by Professor James P. LeSage [Fields Endowed Chair of Urban and Regional Economics, Department of Finance and Economics, Texas State University - San Marcos]:

Topic: Spatial econometric panel data model specification: A Bayesian approach(in English)

For the after-talk casual meeting (drinks, finger-food & Xmas cookies) at the Department Lounge (Room D4.3.210) please register via Email to wgi-team@wu.ac.at by Monday, Dec. 16, 2013, 10 AM.
We hope to see you all there!
 

Abstract. Taking a Bayesian perspective on model uncertainty for static panel data models proposed in the spatial econometrics literature considerably simplifies the task of selecting an appropriate model. A wide variety of alternative specifications that include various combinations spatial dependence in lagged values of the dependent variable, spatial lags of the explanatory variables, as well as dependence in the model disturbances have been the focus of a literature on various statistical tests for distinguishing between these numerous specifications.

A Bayesian model uncertainty argument is advanced that logically implies we can simplify this task by focusing on only two model specifications. One of these, labeled the spatial Durbin model (SDM) implies global spatial spillovers, while the second, labeled a spatial Durbin error model (SDEM) leads to local spatial spillovers. A spatial spillover arises when a causal relationship between characteristics/actions of entity/agent (Xi) located at position i in space exerts a significant influence on the outcomes/decisions/actions (Yj) of an agent/entity located at position j. Formally, LeSage and Pace (2009) define this as: ϑYj = ϑXi = 0. If locations j are neighbors to location i, we have a local spatial spillover. In contrast, if locations j include not only neighbors to i, but neighbors to neighbors of i, neighbors to neighbors to neighbors, and so on, we have a global spillover.

A Bayesian approach to determining an appropriate local or global specification, SDEM versus SDM is set forth here for static panel variants of these two models. The logic of the Bayesian view of model uncertainty suggests these are the only two specifications that need be considered. This greatly simplifies the task confronting practitioners when using static panel data models.

Keywords: Static space-time panel data models, Bayes factors, local versus global spatial spillovers

Download paper from SSRN

For more information see the personal Home Page of James P. LeSage, James P. LeSage at IDEAS, and learn about his Econometrics Toolbox for MATLAB